> For the complete documentation index, see [llms.txt](https://tonify-1.gitbook.io/tonify-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://tonify-1.gitbook.io/tonify-docs/policy/what-next.md).

# What to do

An acceptable formula reform — and why royalties-for-plays should still be abandoned. Policy reading, not a theorem.

**The level.** This page is a policy reading of the study, not a fifth lemma. It sits on L1–L4, sim1–sim6 and emp2. If it diverges from them, it is wrong.

**The decision.** Fix the pool formula if you must stay in World A — four steps, all computed. Change the cash register anyway: the unit of value is the fan, not the play. What used to be a royalty becomes the price of a thing — a drop, a ticket, patronage. Tonify is the wedge; Unify is the rail.

***

### 1. The acceptable reform (World A)

The industry has spent a decade arguing about the pool formula. The argument is not stupid: the formula is the only lever the platform holds. The contract is held by the label. Finding 9 is the same fact: the platform treasury is nearly indifferent to the artist's contract (−0.19%), and the platform still holds the pen.

If you stay in the pool, an acceptable reform is four steps — all four already computed.

**1. The contract outweighs the rule.** Signed → independent moves MVA by ×14.8; pro-rata → user-centric moves it by ×1.34 at the baseline wallet, and that is not even a scalar: the effect flips sign at u\* ≈ 14,146 plays/year (L2; fig14; pass-through grid fig19: even ρ = 20% is ×5.0 against ×1.34). World A's best formula does not compensate for a 6.8% label pass-through: user-centric signed (140,463) is worse than pro-rata independent (12,771). Reforming the rule without reforming the contract is a reshuffle one order of magnitude short.

**2. User-centric is not “fairer to the niche.”** L1: UC/PR is the mean of P̄/Pᵤ over the artist's own audience. What wins is lightness and heterogeneity of listeners, not nicheness. At the measured coupling γ ≈ +0.12, user-centric takes from the head four times harder than independence predicts (top 0.28%: 33.9% → 8.9%). Apply it only where the audience sits below u\*. Above u\*, the reform *hurts* that artist.

**3. Fraud detection.** fig3: injecting F% bot streams siphons F/(1+F) from everyone. At 30% the pool loses 23%. In the direct economy a bot cannot donate other people's money; own losses (chargebacks) do not smear.

**4. The ceiling of any stable formula is known in advance.** Bergantiños & Moreno-Ternero: any stable rule divides a listener's fee only among the artists that listener played. An artist's ceiling in World A is their own audience, under any formula. The formula reshuffles money inside that audience. It does not create a listener.

That is the maximum of World A. It is acceptable hygiene of the pool. It is not enough, because the unit of account remains the play.

***

### 2. Why royalties-for-plays should be abandoned

A streaming royalty is a residual on a play in a shared (or per-wallet) pool. The cash register is the play. In that ledger a fan is indistinguishable from background listening, a bot, and an AI track. Hence:

* 87% of tracks never cross 1,000 streams/year (Luminate; anchor T1).
* 89.9% of signed artists wait more than ten years for a first payout at a $13 threshold.
* 75% of artists earn exactly zero from direct payments at the calibrated superfan share (L3, sim4-D); among artists with $A\_i>0$, 73.3% receive zero from direct — and that is *after* the switch. The direct economy does not save the tail. It relocates inequality: the tail index is unchanged, a zero atom appears at the bottom.
* L4: the zero atom blocks stochastic dominance in both directions. Direct is not unconditionally “better than the pool” at any k. Raising payment frequency does not kill the lottery. Conversion σ does; cadence does not.
* The operator that attacks the atom is not a fourth pure mechanism. **Direct + guaranteed discovery floor** keeps fan money direct and adds a platform reserve $B$ for artists with verified audience $A\_i\ge A\_{\min}$, weighted $A\_i^\beta$ with $\beta<1$. Any $B>0$ yields $q\_{\mathrm{eligible}}=0$. On the sim1 world the atom is an eligibility cost ($A\_{\min}=1$ kills it among $A\_i>0$; $A\_{\min}\le 3$ leaves at most half); a living wage is a $B$ cost (5% of the independent pool is $16.38/year mean floor at $A\_{\min}=10$, and hybrid MVA barely moves). This is a hybrid: $B$ has to come from somewhere. That is a feature. sim6.

The unit of value is not the play. The unit of value is the fan.

That is not a slogan. It is measured in an adjacent industry: a Twitch motivation model (N=1,097) explains hours and free follows and only 3.7% of the variance of *paid* subscriptions; of five motives, payment is predicted by one — social-relational. People do not pay for hours. They pay for a tie. The pool cannot see a tie: under pro-rata a fan and a background play are the same object; under user-centric they differ only by wallet intensity, not by willingness to pay.

The order of one action, not a career forecast: a $5 donation ≈ 1,667 streams at the shop-window $0.003/stream and ≈ 16,700 at a signed artist's pocket ($0.0003). The TON rail delivers 94.9¢ of a dollar, Stars mobile 64.1¢. This compares a voluntary payment with a passive stream. What share of listeners pay at all is an open dial (emp2: k UNMEASURED; σ is the 0.6–1.7% band).

So abandoning royalties-for-plays is not a moral. It is a change of cash register. The play remains a discovery surface (sim2: hubs at a small budget ignite a cascade, random does not; p\* ≈ 0.15 is the product's K-factor). The cash register is a transaction from a specific person to a specific artist.

***

### 3. What a royalty becomes

The word does not have to die. It has to stop being a *residual on a flow* and become a *price of a thing*.

In World A a royalty is a share of a pool, smeared across plays. In World B the same word means a discrete transfer of an object:

* sale of the recording as a digital asset (drop, NFT, digital album) — a number, not a share;
* a ticket (live or virtual) — a number;
* patronage / donation / artist subscription — a number, with rate k and check c.

This is not “tokens instead of labels.” It is a return to the accounting music knew before the pool: a thing has a price, the price goes to the author, a play by itself accrues nothing. NFT here is not an aesthetic; it is a form: a unique object with a recorded transfer, not a smeared residual. A ticket is the same form without a chain. The difference from the pool is one unit: the play is no longer the cash register.

The honest promise of this register is not “everyone gets rich.” sim3: 9,010 of 10,000 artists exit within 36 months even when the treasury law holds and the money is real. No formula saves everyone — nine in ten leave because they are not listened to. L3 says the same in the language of tails: the index does not get lighter. The promise is different: the artist who is listened to stops counting decades to a first payout; the tie to their people belongs to them, not rented from a platform; of a fan's dollar, 80–95¢ arrives, not 6¢ after the label.

“One thousand true fans” is empirically beaten (Death by 1000 true fans; a power law on Patreon). We do not promise a thousand. We promise that one person who cares is worth orders of magnitude more than one play. That is a claim about a unit, not about everyone.

***

### 4. Tonify is the wedge. Unify is the rail.

Tonify is not another player. It is a coordination layer sewn into a behaviour that already exists: a track in a Telegram chat. The wedge is behavioral, not technological. CAC ≈ 0 by construction (a share is distribution is acquisition). Shadow profiles break catalogue cold-start: the artist arrives to an audience already built. The treasury law — payouts ≤ external inflow — is what separates this from the emission graveyard (sim3: 200/200 emission regimes lose ≥80% of peak; the law cannot bankrupt its treasury and does not promise product immortality).

Unify is what Tonify becomes if the loop proves itself on external money. TCP/IP never owned content — it owned how content moves. Unify does not own music — it owns how music is monetised: a layer of identity, payments and rights-clearing that a messenger, a game, a car, a pair of headphones can plug into. GTA still plays fake music in fake cars not because Rockstar does not want the real Carti, but because every song has two rights held by different people, licences expire, and there is no such thing as an “album licence.” One connect instead of a hundred bilateral deals is the rail. Stripe for payments, Plaid for bank data, Unify for music value.

The cautionary tale is RSS: a protocol that worked died because walled gardens paid creators more. A rail wins only if it pays artists better than any garden. That is the entire design: a 5% take against the Twitch-mechanics 50/50 split and against the 6.8% signed-pool pocket.

The chain the study can already compute, and the product still has to measure:

```
cascade (sim2) → superfan conversion σ (open) → k, check (emp2, UNMEASURED) → treasury (sim3)
```

sim5 computes it as a map. Until the pilot CSV is full, the map is not a fact.

***

### 5. What this does not promise, and how it dies

* **Not a theorem.** L1–L4 are about a fixed play matrix. This page is a decision to change the cash register. Equilibrium (the matrix itself moves when the rule moves) is open in the study.
* **Not a $75B forecast.** An illustration of direction from the manifesto: where the pie flows when the middleman is removed. Removing the middleman does not, by itself, bake a second pie.
* **Not “World B is mathematically inevitable.”** Certainties do not need founders. This is a bet. Realistically most such bets die; the upside is not incremental if incentives align at scale.
* **k and σ are unmeasured in Telegram music.** Adjacent industries are colour, not a closer (emp2). If measured k < 0.38, direct loses to the independent pool on every remaining axis. If σ stays near zero, L4: no k closes the tail lottery.
* **9,010 / 10,000.** A direct economy does not mint a listener from air.
* **The room is rented.** Stars on a phone clear through Apple's till; “we remove the middleman” is false on mobile today. The TON rail is the answer, not a slogan.

The falsifier of this whole page, in one sentence: if a Telegram fan does not pay an artist more often, and does not convert more often, than the worst corner of 0.38–6.31 × 0.6% × $3.10 — World B loses to World A on this market, and the honest formula reform (§1) remains the ceiling.

Until then the decision is this. Fix the formula — yes, in the four steps above, if you are a platform inside the pool. Build the cash register on the fan, not on the play. Take royalties-for-plays off the register and leave the play as discovery. Turn what used to be a royalty into the price of a thing: a drop, a ticket, patronage. Tonify proves that in the one window where the behaviour already exists. Unify carries it to where music plays and a cash register does not yet.

*MIT. This page is not executed by code — it is read against L1–L4, sim1–sim6 and emp2. If it diverges from them, it is wrong, not they.*


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://tonify-1.gitbook.io/tonify-docs/policy/what-next.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
